linux-kernel - Re: [PATCH v3] mm: fix tick timer stall during deferred page init

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <20200401171331.26v6th25ezhixwpt@ca-dmjordan1.us.oracle.com>
Date:   Wed, 1 Apr 2020 13:13:31 -0400
From:   Daniel Jordan <daniel.m.jordan@...cle.com>
To:     Pavel Tatashin <pasha.tatashin@...een.com>
Cc:     Michal Hocko <mhocko@...nel.org>,
        Daniel Jordan <daniel.m.jordan@...cle.com>,
        Vlastimil Babka <vbabka@...e.cz>,
        David Hildenbrand <david@...hat.com>,
        Shile Zhang <shile.zhang@...ux.alibaba.com>,
        Andrew Morton <akpm@...ux-foundation.org>,
        Kirill Tkhai <ktkhai@...tuozzo.com>,
        linux-mm <linux-mm@...ck.org>,
        LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v3] mm: fix tick timer stall during deferred page init

On Wed, Apr 01, 2020 at 12:51:56PM -0400, Pavel Tatashin wrote:
> > > I do not remember seeing any real failures, this was a theoretical
> > > window. So, we could potentially simply remove these locks until we
> > > see a real boot failure in some interrupt thread. The allocation has
> > > to be rather large as well.
> >
> > Yes please! We are really great at over complicating and over
> > engineering stuff based on theoretical issues and build on top of that
> > and make the code even more complex because nobody dares to re-evaluate
> > and so on.
> 
> I will submit a patch (or revert) whichever is cleaner.

I had thought people would be concerned about that window.  I believe the quote
from the time it was being discussed was "rare failures suck."  They very much
do :) but in this case I think any failure would be relatively easy to
diagnose.

So great, simpler is always better, and I'll wait for what you send, Pasha.

Alternatively, if you won't be able to get to it for a while, I can write it
up, having thoroughly paged all this in over the past week.